摘要
SPAD值是评价烟草健康状况的重要农学参数。本文利用ASD地物光谱仪和叶绿素仪(SPAD-502)连续两年测定了TMV胁迫下烟草冠层的光谱数据和SPAD值。采用减量精细采样法,筛选了TMV胁迫下烟草SPAD值的光谱敏感波段,并构建了预测模型。结果表明:TMV胁迫下烟草SPAD值的敏感波段主要位于可见光区域及近红外区域,其最佳波段为731 nm、741 nm、835 nm和910 nm。构建的SPAD值的差值植被指数DVI(R835,R910)、比值植被指数RVI(R731,R741)、归一化植被指数NDVI(R731,R741)线性模型、逐步回归(SMLR)和BP神经网络等预测模型,决定系数(R^(2))分别为0.62、0.60、0.61、0.69和0.89;模型的测试精度(P-R^(2))分别为0.49、0.52、0.52、0.51和0.79,均方根误差RMSE分别为5.33、5.22、5.23、17.13和3.40。表明可利用DVI(R835,R910)、RVI(R731,R741)、NDVI(R731,R741)和BP神经网络来预测烟草SPAD值,并进一步监测其健康状况。
The SPAD value is an important parameter for assessing tobacco health.The canopy hyperspectral reflectance and chlorophyll content(SPAD)of tobacco under tobacco mosaic virus(TMV)stress were measured by the ASD portable spectrometer and portable chlorophyll meter(SPAD-502)for consecutive two years.The reduced precise sampling method was adopted to select the spectral sensitive bands for SPAD values of tobacco infected by TMV,and the estimating model of SPAD values of tobacco was built.The results showed that the sensitive bands for SPAD values of tobacco infected by TMV were concentrated in the visible range and near-infrared range,and the optimum band was 731 nm,741 nm,835 nm and 910 nm,respectively.The DVI(R835,R910),RVI(R731,R741),NDVI(R731,R741),SMLR and BP neural network models for SPAD values had R^(2) values of 0.62,0.60,0.61,0.69 and 0.89,P-R^(2) values of 0.49,0.52,0.52,0.51 and 0.79,RMSE values of 5.33,5.22,5.23,17.13 and 3.40,respectively.The results indicated that DVI(R835,R910),RVI(R731,R741),NDVI(R731,R741)and BP neural network could be used to predict the SPAD value of tobacco and monitor its healthy state.
作者
许跃奇
阎海涛
常栋
王晓强
何晓冰
马文辉
贾方方
XU Yue-qi;YAN Hai-tao;CHANG Dong;WANG Xiao-qiang;HE Xiao-bing;MA Wen-hui;JIA Fang-fang(Pingdingshan Tobacco Company of Henan Province,Pingdingshan 467000,China;College of Biology and Food,Shangqiu Normal University,Shangqiu 476000,China)
出处
《江西农业学报》
CAS
2022年第3期8-15,共8页
Acta Agriculturae Jiangxi
基金
河南省青年人才托举工程资助项目(2020HYTP024)
河南省烟草公司平顶山市公司重点项目(PYKJ202101)。